About Statistics



A great movie called Moneyball is really all about baseball, a true story about two guys Bille Bean and Peter Brand, who developed a system for selecting a winning team that cost substantially less than millonaire teams like the New York Yankees would spend. The thing behind this story was that not only did they understand that some players were undervalued and bought these players, but they had the courage to stick with their system in the face of a string of successive losses, despite the world thinking they were crazy.

It’s much like trading in that sence, as once you have a system that you have devised and backtested, you have to have belief in it and trade it, even when you have those periods of losses. Armed with the statistical knowledge, the trader might gain additional confidence in his trade idea and might even trade it more aggressively. Knowing what to expect based on precedent—and then seeing if the market actually follows its historical tendencies—helps prepare the minds of discretionary traders for a variety of market scenarios.

As example, I know from my statistical studies that 70% of the time, the balanced markets tend to revert back to their value (BTV), especially under conditions of relatively low momentum. Another stat shows that 75% of the time, the imbalanced markets tend to seek balanced targets (PBT), especially under conditions of relatively high momentum.

Another study shows that if a market opens around previous value area and then remains for two half-hour periods into value, there is an 70% chance that market will rotate all the way to the other side of value (VAR). One more stat shows that if a market prints a new record high at ETH, there's 80% chance that record high can be revisited during RTH.

These are some statistical E-Mini SP500 numbers, that can provide guidelines for discretionary traders:

-80% of the time, one side of overnight (ON) range is tested.
-75% of the time, one side of prior day range (PD) is broken.
-78% of the time, one side of initial balance (IB) is broken.
-65% of the time, the average gap (6p) has been closed.
-80% of the time, the half gap (50% RTH gap) is tested.
-18% of the RTH days are Trend days (Multimodal Profiles).
-72% of the RTH days are Normal days (Bimodal/Trimodal).
-10% of the RTH days are Neutral days (Inside ON Range).



Remember that context is extremely important, don't try to use any statistic as a rigid, systematic trading rule, blindly applied to all market conditions. Statistical analysis provides an hypothesis only; current market action will either support or refute these hypothesis.